Original article PROPIONIBACTERIUM ACNES PHYLOGENETIC TYPE III IS ASSOCIATED WITH ...

23 Pages·2017·0.61 MB·English

European Journal of Microbiology and Immunology 7 (2017) 1, pp. 37–45
Original article
DOI: 10.1556/1886.2016.00040
PROPIONIBACTERIUM ACNES PHYLOGENETIC TYPE III IS ASSOCIATED
WITH PROGRESSIVE MACULAR HYPOMELANOSIS
Rolf L. W. Petersen, Christian F. P. Scholz, Anders Jensen, Holger Brüggemann, Hans B. Lomholt*
Department of Biomedicine, Aarhus University, Aarhus, Denmark
Received: December 4, 2016; Accepted: December 21, 2016
Progressive macular hypomelanosis (PMH) is a skin disorder that is characterized by hypopigmented macules and usually seen in
young adults. The skin microbiota, in particular the bacterium Propionibacterium acnes, is suggested to play a role.
Here, we compared the P. acnes population of 24 PMH lesions from eight patients with corresponding nonlesional skin of the
patients and matching control samples from eight healthy individuals using an unbiased, culture-independent next-generation
sequencing approach. We also compared the P. acnes population before and after treatment with a combination of lymecycline and
benzoylperoxide.
We found an association of one subtype of P. acnes, type III, with PMH. This type was predominant in all PMH lesions (73.9%
of reads in average) but only detected as a minor proportion in matching control samples of healthy individuals (14.2% of reads in
average). Strikingly, successful PMH treatment is able to alter the composition of the P. acnes population by substantially diminish-
ing the proportion of P. acnes type III.
Our study suggests that P. acnes type III may play a role in the formation of PMH. Furthermore, it sheds light on substantial
differences in the P. acnes phylotype distribution between the upper and lower back and abdomen in healthy individuals.
Keywords: progressive macular hypomelanosis, Propionibacterium acnes, Cutibacterium acnes, next-generation sequencing,
subtype III, skin microbiota, single locus sequencing type, phylotype
Abbreviations: PMH, progressive macular hypomelanosis; SLST, single-locus sequencing typing; ST, sequence type; NGS, next-
generation sequencing; BPO, benzoylperoxide
Introduction Several treatment modalities are used against PMH
including topical benzoylperoxide 5% (BPO) and
Progressive macular hypomelanosis (PMH) is character- clindamycin 1% alone or in combination with ultraviolet
ized by symmetric nonscaly hypopigmented skin areas A (UVA) or narrow band ultraviolet B (UVB) irradia-
that are predominantly visible in the sebaceous areas of the tion [4–7], oral lymecycline in combination with topical
trunk. In the lower back and abdomen, discrete lesions are BPO 5% for 3 months [8], and low-dose isotretinoin for
distinguished while they are more confl uent on the upper 1 month [9]; however, the ideal treatment is not yet de-
trunk. There is no infl ammation, pain, or itching associated fi ned.
with PMH, but the disease can have a major psychosocial The etiology of PMH is not known; however, several
effect on patients. PMH appears to be more frequent in studies indicate that the Gram-positive anaerobic bac-
young women, and although the disorder has a worldwide terium Propionibacterium acnes may play a pivotal role
distribution, it is most often identifi ed in dark-skinned [10, 11]. Biopsy specimens from PMH lesions contained
populations [1–3]. P. acnes in pilosebaceous ducts in contrast to biopsies
* Corresponding author: Hans Bredsted Lomholt; Department of Biomedicine, Aarhus University, Wilhelm Meyers Allé 4,
8000 Aarhus C, Denmark; Phone: [phone]; E-mail: [email protected]
This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 License, which permits unrestricted
use, distribution, and reproduction in any medium for non-commercial purposes, provided the original author and source are credited.
First published online: February 27, 2017 ISSN 2062-8633 © 2017 The Author(s)
38 R. L. W. Petersen et al.
from healthy skin [11], real-time PCR showed a signifi cant Sampling sites and procedure
predominance of P. acnes in lesional skin as compared to
nonlesional skin [10], and red fl uorescence was detected Samples were taken in all patients from three lesional ar-
in lesions when subjected to Woods light [11]. Finally, eas including the lower, middle, or upper back (depending
antibacterial treatment effective against P. acnes leads to on the position of the PMH lesions) and the abdomen, and
repigmentation [4, 8]. It is not known how P. acnes could additional samples were taken from patients in unaffected
induce hypopigmentation. Using microscopy, a decrease adjacent skin areas. One patient had no lesions on the ab-
in melanin production and a change in the distribution of domen and was sampled from all three positions on the
melanosomes with a resultant decrease in epidermal mela- back. Six of the patients were additionally sampled after
nin was shown in PMH lesions [12, 13]. treatment from the same areas as the pretreatment sample.
Based on multilocus sequence typing (MLST) and sin- Samples from the lower and upper back, and abdomen were
gle-locus sequence typing (SLST) schemes and complete taken from the eight healthy controls. In addition, samples
genome sequencing, the population of P. acnes has been from the forehead and the buccal cavity were taken from
shown to consist of several phylogenetic subtypes com- patients (Fig. S1). Samples for NGS-based SLST were ob-
monly designated IA1, IA2, IB, IC, II, and III [14–21]. tained by swabbing the skin fi rmly for 20 s with a sterile
A previous study based on bacterial cultivation has in- cotton swab moistened in sampling buffer (0.1% detergent
dicated that certain subtypes of P. acnes may be associ- [Triton X-100] in 0.075 M phosphate buffer, pH 7.9) [25].
ated with PMH: an abundance of a specifi c but unidenti- DNA was extracted from all collected samples, and SLST
fi ed type of P. acnes was observed in PMH lesions that is fragment amplifi cation was performed as described pre-
different from P. acnes types isolated from acne lesions viously [20]. The amplicons were then subjected to next-
[22]. Recently, we cultured P. acnes isolates belonging to generation sequencing (NGS) using the pyrosequencing
the otherwise uncommon type III from lesions of PMH technology. Subsequent bioinformatics analysis included
patients and sequenced their genomes [23], and a very quality control and sequence read assignment to the STs
recent study revealed an abundance of P. acnes type III of the SLST scheme (http://medbac.dk/slst/pacnes), result-
in bacterial cultures from lesional skin in 14 of 34 PMH ing in a high-resolution phylotype analysis of the P. acnes
patients [24]. population in each sample.
In the present study, the type distribution of the entire Samples for bacterial cultivation were taken with a
population of P. acnes in affected and unaffected skin ar- sterile charcoal swab moistened in sampling buffer fi rm-
eas of PMH patients, including samples after treatment, ly scrubbed on the skin for 20 s. From each patient, two
and matching control samples was determined using a swabs were obtained from lesional skin on the back and
culture-independent next-generation sequencing (NGS)- abdomen, respectively, and, in addition, two swabs from
based SLST approach. Results show a strong association adjacent nonlesional skin. For a semiquantitative estima-
of P. acnes type III with disease. tion of the number of bacteria, the primary charcoal cot-
ton swabs were streaked on tryptone-yeast-glucose (TYG)
agar plates and incubated for 120 h in an anaerobic cham-
Materials and methods ber (Forma Scientifi c Anaerobic System model 1024).
Patient and control cohort, treatment regimen
Next-generation sequencing-based single-locus sequence
Eight patients with PMH were recruited by voluntary typing (NGS-based SLST)
consent in a private dermatology practice in Aalborg,
Denmark. The patients were all clinically examined by Cotton swab samples from both patients and healthy con-
a specialist in dermatology (H.B. Lomholt), and PMH trols were transferred to 1.5-ml Eppendorf tubes, and bac-
was diagnosed based on the fi nding of clinically char- terial DNA was isolated and purifi ed using PowerLyzer®
acteristic lesions, patient history, and a lack of Malas- PowerSoil® DNA Isolation Kit (MO BIO, Carlsbad, CA,
sezia in microscopic inspections. All patients were fe- United States) according to the manufacturer’s instruc-
male between 18 and 31 years of age (mean, 23.5 years) tion. The isolated DNA was then subjected to amplifi ca-
(see Supplementary Table S1 online). Five of the patients tion by PCR as described previously [20]. All primer se-
were treated in a 3-month course with oral lymecycline quences are given in Supplementary Table S3 online. All
(300 mg, daily) combined with a daily wash using a 5% PCR reactions were made by mixture of 5 μl DNA sample,
BPO washing gel, and one patient used only the 5% BPO 2.5 μl AccuPrime PCR Buffer II (Invitrogen), 1.5 μl of
wash (see Supplementary Table S2 online). As controls, forward primer, 1.5 μl of reverse primer (10 μM), 0.15 μl
eight healthy volunteers were recruited among students AccuPrime Taq DNA Polymerase High Fidelity (Invitro-
at the University of Aarhus, Denmark. They were all fe- gen), and 14.35 μl PCR-grade water into a total volume of
male between 24 and 31 years of age (mean, 26.5 years). 25 μl. All samples were amplifi ed by an initial denatura-
Information regarding the study participants and treat- tion for 2 min at 94 °C, followed by 35 cycles of 20 s of
ment is summarized in Supplementary Tables S1 and S2 denaturation at 94 °C, 30 s at 55 °C for annealing, and 60 s
online. of extension at 68 °C, ending with a 5-minute step of ex-
European Journal of Microbiology and Immunology
P. acnes type III in macular hypomelanosis 39
tension at 68 °C. Three PCRs per sample were performed culture was then examined; up to 10 random single colo-
and verifi ed on an agarose gel, pooled together, and pu- nies resembling P. acnes were selected and individually
rifi ed using a NucleoSpin Extract Kit (Macherey-Nagel). subcultivated on new TYG agar plates for 72 h under an-
The concentration of purifi ed DNA samples was meas- aerobic conditions. After growth, the P. acnes isolates were
ured with a NanoDrop 2000 spectrophotometer (Thermo harvested and subjected to DNA extraction by a boiling
Scientifi c). The amplicons were pooled in batches of 20 procedure at 100 °C for 10 min in 0.5 ml Eppendorf PCR
and sequenced unidirectionally from the forward primer tubes containing 300 μl PCR-grade water. A PCR was car-
using Roche GS FLX+ pyrosequencing technology either ried out with the SLST primers (see Supplementary Ta-
at Institute of Microbiology and Genetics, Georg-August ble S3 online) as follows: 2 μl of the crude DNA extract
University Göttingen, Göttingen, Germany or at Euro- was mixed with 1 μl of each forward and reverse primer
fi ns Genomics, Ebersberg, Germany. The data were then (10 μM), 10 μl 5´-PRIME Hotmastermix (5 PRIME, Ham-
processed using the PyroNoise implementation in Mothur burg, Germany), and 11 μl of PCR-grade water. PCR con-
v. 1.36.1 [26]. All sequences have been deposited at NCBI ditions were as follows: initial denaturation of 40 s at 96 °C
with the project number PRJNA347641. The sequence followed by 35 cycles of 35 s of denaturation at 96 °C, 40 s
reads were aligned to all the known STs in the SLST da- of annealing at 55 °C, and 40 s of extension at 72 °C, fol-
tabase (http://medbac.dk/slst/pacnes) using BLASTn [27] lowed by a fi nale 7-minute extension step at 72 °C. The
and assigned an individual ST based on a best-hit model resulting PCR products were run on a 1% agarose gel to en-
with a cut-off value of 99.5%; anything below this thresh- sure quality and were then sequenced at GATC Biotech AG
old or reads with two or more identical best hits were dis- (Konstanz, Germany) with the forward and reverse SLST
carded as unassigned reads. primers. The sequences were assembled and trimmed to the
known SLST fragment size (484 bp) using MEGA v.6.06
[28]. Finally, the resulting sequences were assigned to STs
Sanger sequencing-based SLST of bacterial isolates using the SLST database (http://medbac.dk/slst/pacnes).
SLST typing of P. acnes isolates has been described pre-
viously [20]. In brief, the sampled charcoal cotton swabs Statistical analysis
were streaked on tryptone-yeast-glucose (TYG) agar plates
and incubated for 72 h in an anaerobic chamber (Forma The proportions of P. acnes type III were compared be-
Scientifi c Anaerobic System model 1024). This primary tween mean values of the three lesional sites in patients
Fig. 1. Comparison of P. acnes ST distribution in PMH samples and controls based on next-generation sequencing data. a) Average of
P. acnes ST proportions in patient versus controls samples. b) Each column represents ST proportions as an average of the three sam-
pling sites (upper back, lower back, and abdomen). Data are given for eight patients, P1 to P8, and eight controls, C1 to C8. Each ST
(A to L) is given a color as indicated, and the corresponding P. acnes subtype is given in brackets
European Journal of Microbiology and Immunology
40 R. L. W. Petersen et al.
and corresponding sites in controls using the unpaired Wil- a higher proportion of type IA1 (STs “A” to “E”) strains
coxon rank sum test/Mann–Whitney U test. Lesional sites (56.8%) and type II (ST “K”) strains (22.3%), while these
in patients before and after treatment were compared using were found in much lower proportions in PMH lesions:
the paired Wilcoxon signed rank test. type IA1 (12.8%) and type II (5.6%). Individual variation
among the patients and controls was observed: the pro-
portion of type III in the total P. acnes population varied
Ethics statement between 50% and 91% in PMH lesions and between 0%
and 53% in control samples (Fig. 1b).
The study protocol was approved by the Ethics Commit- We wanted to confi rm these fi ndings with a culture-de-
tee of Region North, Denmark (document N-20120050), pendent technique: swab samples taken from PMH lesions
and the study was conducted according to the principles of and healthy controls were cultivated anaerobically. Up to 10
the declaration of Helsinki. Written informed consent was P. acnes colonies per sample were randomly selected from
obtained from all study participants. the agar plates; for each isolate, the classical SLST assign-
ment by Sanger sequencing of the PCR-amplifi ed SLST
fragment was carried out. In average, 39.6% of the P. acnes
Results colonies were type III in lesion samples, in contrast to only
12.5 % in controls (see Supplementary Fig. S2 online).
Type III P. acnes predominates in PMH samples
SLST amplicons from 96 samples were selected for pyro- P. acnes type III in PMH patients is enriched in PMH
sequencing. In average, 7911 sequence reads per sample regions and adjacent skin areas but is rarely found at
were obtained and 97% of the reads could be assigned to a other body sites
P. acnes sequence type (ST) (see Supplementary Table S4
online). In the whole data set, the SLST scheme distin- A recent study [29] suggested that the phylotype distribu-
guished 90 different STs of P. acnes. tion of P. acnes is relatively uniform and stable throughout
The results of the P. acnes ST distribution in PMH the body. Thus, we wanted to investigate if PMH patients
lesions from eight patients and eight matching healthy are also predominantly colonized by P. acnes type III at
controls are shown in (Fig. 1a,b). A clear distinction was different body sites other than the PMH lesions. Samples
found when comparing PMH lesional samples to matching from different locations including the lower and upper
controls: P. acnes type III (the corresponding ST is des- back, the forehead and the buccal mucosa were analyzed
ignated “L”) was the dominating phylotype in PMH le- from patients and controls.
sions (Fig. 1a); on average, 73.9% of the reads belonged Type III was rarely detected on the forehead or in the
to this type. In contrast, in the matching control samples, buccal mucosa of patients (Fig. 2a). Most patients had a
only 14.2% of the reads could be assigned to P. acnes type mixture of different P. acnes types on the forehead, in par-
III (p value 7.8 × 10−5). Instead, control samples contained ticular, strains of the phylotypes IA2, IC, and II. Only one
Fig. 2. Average P. acnes ST distribution shown for different body sites in patients and controls. a) P. acnes ST distribution in patient
PMH lesions on lower back, upper back, and abdomen, and unaffected skin on the forehead and buccal mucosa. b) ST distribution in
controls at sites corresponding to patient lesions on lower back, upper back, and abdomen. Each ST (A to L) is given a color as indi-
cated, and the corresponding P. acnes subtype is given in brackets
European Journal of Microbiology and Immunology
P. acnes type III in macular hypomelanosis 41
Fig. 3. Comparison of average P. acnes ST distribution in PMH patients in lesional versus nonlesional skin based on next-generation
sequencing data. Each ST (A to L) is given a color as indicated, and the corresponding P. acnes phylobtype is given in brackets
patient was found to have a signifi cant proportion of type dominant P. acnes type on the lower back skin of healthy
III P. acnes on the forehead. On the buccal mucosa, type controls was type II (33.5%), a type that was rarely de-
IA1 was predominant in most patients. tected on the upper back (4.0%). The dominating type of
Looking at the P. acnes phylotype distribution of pa- the upper back and the abdomen was type IA1 (44.7% and
tients on nonlesional skin sites, a dominance of type III 40.6%, respectively).
strains was detected, albeit at a lesser extent than on le-
sional skin, 53% versus 74%, respectively (Fig. 3).
Healthy individuals were analyzed as well. Interesting- PMH treatment alters the P. acnes population
ly, in average, we could detect a larger proportion of type and diminishes the proportion of type III strains
III strains (21.7%) on the lower back skin compared to
the upper back skin (5.0%), indicating that the lower back Next, we wanted to investigate how the treatment of PMH
is the preferred habitat for type III strains (Fig. 2b). The with a combination of lymecycline and BPO might alter
Fig. 4. Clinical PMH treatment responses. Lesional skin on the back before and after treatment with lyme-
cycline 300 mg and BPO daily washes for 3 months shown for patient P2 (A and B) and patient P4 (C and D)
European Journal of Microbiology and Immunology
42 R. L. W. Petersen et al.
Fig. 5. Comparison of P. acnes ST distribution in PMH patients before and after treatment based on next-generation sequencing data.
a) Average P. acnes ST proportions in patient samples before and after treatment. b) Each column represents ST proportions as an
average of the three sampling sites (upper back, lower back, and abdomen). Data are given for six patients before and after treatment.
Each ST type A to L is given a color as indicated, and the corresponding P. acnes subtype is given in brackets
the P. acnes type distribution. Six patients were treated for ment from the same skin location. Information about the
3 months, and samples were taken before and after treat- treatment regimen and the response for each patient is
Fig. 6. Comparison of average ST distribution among P. acnes subtype III strains in PMH patients versus
controls based on next-generation sequencing data. Each type III ST (L1 to L6) is given a color as indicated
European Journal of Microbiology and Immunology
P. acnes type III in macular hypomelanosis 43
given in Supplementary Table S2 online. Representative ruled out. Therefore, up to 10 randomly selected colonies
images of the back skin of two patients who responded were cultured from samples and analyzed by traditional
well to the treatment are shown (Fig. 4). Sanger sequencing for comparison. Importantly, the domi-
We could detect a striking reduction of the proportion nance of type III strains in PMH lesions as compared to
of type III P. acnes from an average of 80% before treat- controls was consistent in both techniques.
ment to 22% after treatment (p = 0.015) (Fig. 5a). In all P. acnes subtype III was fi rst reported as a new phy-
six patients, the type III proportion was diminished after logenetic type in 2008 based on four strains isolated from
treatment in PMH-affected skin sites. In three patients spinal intervertebral disc material [17]. In addition, sub-
(P2, P4, and P8), the type III population was almost com- type III isolates were recently detected in surgically ex-
pletely eradicated after treatment (Fig. 5b). Interestingly, cised lumbar disc herniations from 5 of 24 patients [30].
these three patients showed a particular good response Subtype III bacterial cells differ from other P. acnes types
to the treatment with almost no remaining PMH lesions. in showing a long fi lamentous morphology reminiscent of
The type distribution after treatment resembled in aver- Propionibacterium propionicum. Subtype III differs also in
age the one detected in controls samples (Fig. 1b). In biochemical tests, matrix-assisted laser desorption–ioniza-
contrast, in patients with a less good treatment response, tion time-of-fl ight mass spectrometry (MALDI-TOF MS)
a substantial type III population could be still detected spectra, and genetic markers; thus, it was recently pro-
(Fig. 5b). posed as a new subspecies with the name P. acnes subsp.
elongatum [31]. This subtype has rarely been cultured
from healthy controls, opportunistic infections, or acne
Existence of a specific type III linage associated patients [14, 16, 18, 32]. In contrast, the present study
with PMH lesions? showed a striking predominance of subtype III in PMH
lesional samples from the lower, middle, and upper back,
Since type III strains were also detected, albeit fewer, in and abdomen. This corroborates the previous fi nding of a
healthy samples, in particular on the lower back skin, we unique, unidentifi ed P. acnes type in PMH patients [22],
wanted to investigate if PMH-associated type III strains and the recent report that type III strains were cultured
belong to a specifi c lineage that is different from health- from lesions of 14 of 34 PMH patients [24].
associated type III strains. Our SLST scheme can differ- The NGS-based SLST approach provided no data on
entiate six STs within the type III lineage. The analyses the presence of other bacterial species or the total bacte-
showed that PMH-associated type III P. acnes belong pre- rial numbers but only on the proportional distribution of
dominantly to the STs L1 (56%) and L6 (25%); the latter P. acnes subtypes. Previous studies have found a highly
ST was detected at lower rates among health-associated signifi cant increase in P. acnes numbers in PMH lesions
type III P. acnes (9%) (Fig. 6). Overall, our data do not re- [11, 24]. In accordance, we could also confi rm that P. acnes
veal a clear-cut difference between the type III populations is more abundant in lesions compared to nonlesional skin:
of healthy and PMH-affected individuals. a semiquantitative analysis revealed that all PMH samples
from patients had a higher colony-forming unit count as
compared to adjacent nonlesional skin (data not shown).
Discussion As in other studies, no Malassezia fungus was detected
and only few bacteria of other species, mainly Staphylo-
We report the hitherto most detailed data on the distribu- coccus epidermidis, were found.
tion of P. acnes subtypes on multiple skin sites on eight Our study revealed a high proportion (74%) of P. acnes
PMH patients and eight healthy controls, using a highly type III in lesional skin and also a relatively high propor-
discriminative NGS-based SLST approach. The study re- tion (53%) of type III isolates in adjacent nonlesional skin
vealed an association of one particular type of P. acnes, of the patients. In contrast, Barnard et al. cultured P. acnes
the phylotype III, with the skin disorder progressive macu- from nonlesional skin in only one of 34 patient samples,
lar hypomelanosis. Moreover, an indication that subtype indicative of low P. acnes numbers in nonlesional skin.
III is involved in the PMH disease pathogenesis was re- Therefore, the type distribution at normal skin sites could
vealed by the comparison of patient samples before and not be assessed in their study [24]. In addition, the differ-
after treatment: therapy with lymecycline and BPO, both ence may refl ect different sampling and sample processing
highly active against P. acnes, led to a diminished propor- approaches in the two studies. Barnard et al. cultivated bac-
tion of type III, which was paralleled by the disappearance teria from a homogenized 4-mm punch skin biopsy sample,
of clinical PMH lesions. whereas we used surface skin swabs from a circular area
The applied NGS-based SLST approach gave an aver- of approximately 1.5 cm in diameter. We noticed that it is
age of 7911 reads per sample with 97% of reads assigned diffi cult to completely separate lesions from adjacent non-
to known P. acnes STs. This provided a robust basis for lesional skin sites as lesions may be small and plenty and
a high-resolution estimate of the type distribution in each not well defi ned. This may explain the relatively high pro-
sample. Though this technique is regarded as less biased portion of type III strains in nonlesional skin of patients in
than culture-dependent techniques, a PCR bias due to dis- the present study. Normal controls harbored only a minor
proportional amplifi cation of certain sequences cannot be proportion of P. acnes type III on corresponding skin sites.
European Journal of Microbiology and Immunology
44 R. L. W. Petersen et al.
The SLST scheme can distinguish six different STs normal human microbiome, P. acnes subtype III appears to
among type III strains and four of these were detected in constitute only a minor portion, mainly residing on the low-
PMH lesions with no clear differences in their relative pro- er back and abdomen, among the P. acnes population. The
portions in patients and controls. Interestingly, a regional fi ndings open for future studies of specifi c type III traits to
difference in prevalent type III clones was suggested, further defi ne the role of the bacterium and increase our
based on comparisons of PMH isolates from Europe and understanding of the PMH disease pathogenesis.
Brazil [24].
The mechanism leading to macular hypomelanosis
Funding sources
is not known, but ultrastructural studies have shown less
melanized and aggregated melanosomes instead of single The work was funded in part by the Danish Medical Re-
mature melanosomes transferred from melanocytes to ke- search council (DFF-1331-00241) to H.B. (http://ufm.dk/)
ratinocytes [12]. The association with P. acnes type III and by Fonden for Faglig Udvikling af Speciallægepraksis
suggests that a type III-specifi c factor could be involved. to H.B.L. The funders had no role in study design, data
Comprehensive comparison of type III genomes to P. acnes collection and analysis, decision to publish, or preparation
genomes of other subtypes has identifi ed several genomic of the article.
regions specifi c to type III genomes encoding functions
such as type II secretion system, ABC transporter, inosi-
tol transport/modifi cation, gyrase, integrase, transposase, Authors’ contributions
oligopeptide transport, and processing of sugars/amino ac-
ids [24, 33, 34]. In addition, some genes are absent from H.B.L. conceived the study idea, obtained ethical ap-
type III genomes but present in all other types including proval, included and treated the patients, and collected all
hyaluronate lyase, magnesium-chelatase, iron transporter, bacterial samples from patients and controls. R.L.W. pre-
bacteriocin, 3-isopropylmalate dehydrogenase, maltose pared samples for Sanger and next-generation sequencing.
transporter, and periplasmic binding protein. It has to be R.LW., C.F.P.S., and A.J. analyzed the sequencing data.
investigated in the future if and which factors of P. acnes R.L.W., H.B.L., and H.B. interpreted the data and prepared
type III are important in PMH pathogenesis. the draft article. All authors read, corrected, and approved
At present, there is a major interest in defi ning the nor- the article.
mal human skin microbiome as it is considered important
for maintaining healthy skin. Several metagenomic studies
Competing interests
have described the skin microbiome on species level, but it
has become increasingly clear that subtypes within species The authors state no confl ict of interest.
may play important roles. Most previous P. acnes stud-
ies have shown a predominance of type IA1 followed by
type II and IB among skin isolates derived from the face Acknowledgements
or upper back [14, 16, 18]. The present study is the most
detailed study on P. acnes subtype distribution on different The authors thank Andrea Thürmer, University of Göttin-
body sites. It was shown that subtype III is not normally gen Genomics Laboratory, Göttingen, Germany, for NGS
dominant in skin areas of healthy skin; it was rarely found sequencing support.
on the face and in the oral cavity. On healthy skin, type III
is more frequent at the lower back and abdomen, together
References
with type II, relative to other body sites.
The present study has some limitations. The number
1. Guillet G, Helenon R, Guillet MH, Gauthier Y, Menard N:
of patients and controls is low, even though many samples
Progressive and confluent hypomelanosis of the melano-
were analyzed from each person. In general, nonlesional
dermic metis. Ann Dermatol Venereol 119, 19–24 (1992)
skin was sampled lateral to the lesional area and, therefore, 2. Guillet G, Helenon R, Gauthier Y, Surleve-Bazeille JE,
not on exactly corresponding skin areas. After treatment, Plantin P, Sassolas B: Progressive macular hypomelanosis
faint residual lesions in three patients were detectable, of the trunk: primary acquired hypopigmentation. J Cutan
even though the clinical response was satisfactory. We do Pathol 15, 286–289 (1988)
not know if the bacteria most important in the disease pro- 3. Relyveld GN, Menke HE, Westerhof W: Progressive macu-
cess reside on the skin or in the follicles and the surface lar hypomelanosis: an overview. Am J Clin Dermatol 8,
13–19 (2007)
swab technique employed may have missed bacteria re-
4. Relyveld GN, Kingswijk MM, Reitsma JB, Menke HE, Bos
siding in follicles. Furthermore, detailed data on the total
JD, Westerhof W: Benzoyl peroxide/clindamycin/UVA is
number of bacteria at each site and the presence and pro-
more effective than fluticasone/UVA in progressive macu-
portions of other microorganisms besides P. acnes were
lar hypomelanosis: a randomized study. J Am Acad Der-
not obtained in this study.
matol 55, 836–843 (2006)
In conclusion, the present study showed that P. acnes 5. Santos JB, Almeida OL, Silva LM, Barreto ER: Efficacy of
phylotype III is associated with PMH lesions and, there- topical combination of benzoyl peroxide 5% and clindamy-
fore, may play a role in the disease pathogenesis. In the cin 1% for the treatment of progressive macular hypomela-
European Journal of Microbiology and Immunology
P. acnes type III in macular hypomelanosis 45
nosis: a randomized, doubleblind, placebo-controlled trial. scheme for mixed populations of Propionibacterium acnes
An Bras Dermatol 86, 50–54 (2011) in vivo. PLoS One 9, e104199 (2014)
6. Sim JH, Lee DJ, Lee JS, Kim YC: Comparison of the clini- 20. Yu Y, Champer J, Garban H, Kim J: Typing of Propioni-
cal efficacy of NBUVB and NBUVB with benzoyl perox- bacterium acnes: a review of methods and comparative
ide/clindamycin in progressive macular hypomelanosis. analysis. Br J Dermatol 172, 1204–1209 (2015)
J Eur Acad Dermatol Venereol 25, 1318–1323 (2011) 21. Relyveld GN, Westerhof W, Woudenberg J, Kingswijk M,
7. Kim MB, Kim GW, Cho HH, Park HJ, Kim HS, Kim SH, Langenberg M, Vandenbroucke-Grauls CM, Bos JD,
Kim BS, Ko HC: Narrowband UVB treatment of progres- Savelkoul PH: Progressive macular hypomelanosis is as-
sive macular hypomelanosis. J Am Acad Dermatol 66, sociated with a putative Propionibacterium species. J In-
598–605 (2012) vest Dermatol 130, 1182–1184 (2010)
8. Cavalcanti SM, Querino MC, Magalhaes V, Franca ER, 22. Petersen R, Lomholt HB, Scholz CF, Bruggemann H: Draft
Magalhaes M, Alencar E: The use of lymecycline and genome sequences of two Propionibacterium acnes strains
benzoyl peroxide for the treatment of progressive macular isolated from progressive macular hypomelanosis lesions
hypomelanosis: a prospective study. An Bras Dermatol 86, of human skin. Genome Announc 3, (2015)
813–814 (2011) 23. Barnard E, Liu J, Yankova E, Cavalcanti SM, Magalhaes
9. Kim YJ, Lee DY, Lee JY, Yoon TY: Progressive macular M, Li H, Patrick S, McDowell A: Strains of the Propioni-
hypomelanosis showing excellent response to oral isotreti- bacterium acnes type III lineage are associated with the
noin. J Dermatol 39, 937–938 (2012) skin condition progressive macular hypomelanosis. Sci
10. Cavalcanti SM, de Franca ER, Lins AK, Magalhaes M, de Rep 6, 31968 (2016)
Alencar ER, Magalhaes V: Investigation of Propionibacte- 24. Williamson P, Kligman AM: A new method for the quanti-
rium acnes in progressive macular hypomelanosis using tative investigation of cutaneous bacteria. J Invest Derma-
real-time PCR and culture. Int J Dermatol 50, 1347–1352 tol 45, 498–503 (1965)
(2011) 25. Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M,
11. Westerhof W, Relyveld GN, Kingswijk MM, de Man P, Hollister EB, Lesniewski RA, Oakley BB, Parks DH, Rob-
Menke HE: Propionibacterium acnes and the pathogenesis inson CJ, Sahl JW, Stres B, Thallinger GG, Van Horn DJ,
of progressive macular hypomelanosis. Arch Dermatol Weber CF: Introducing mothur: open-source, platform-
140, 210–214 (2004) independent, community-supported software for describ-
12. Relyveld GN, Dingemans KP, Menke HE, Bos JD, Wester- ing and comparing microbial communities. Appl Environ
hof W: Ultrastructural findings in progressive macular hy- Microbiol 75, 7537–7541 (2009)
pomelanosis indicate decreased melanin production. J Eur 26. Camacho C, Coulouris G, Avagyan V, Ma N, Papadopoulos
Acad Dermatol Venereol 22, 568–574 (2008) J, Bealer K, Madden TL: BLAST+: architecture and appli-
13. Wu XG, Xu AE, Song XZ, Zheng JH, Wang P, Shen H: Clin- cations. BMC Bioinformatics 10, 421 (2009)
ical, pathologic, and ultrastructural studies of progressive 27. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S:
macular hypomelanosis. Int J Dermatol 49, 1127–1132 (2010) MEGA6: Molecular Evolutionary Genetics Analysis ver-
14. Fitz-Gibbon S, Tomida S, Chiu BH, Nguyen L, Du C, Liu sion 6.0. Mol Biol Evol 30, 2725–2729 (2013)
M, Elashoff D, Erfe MC, Loncaric A, Kim J, Modlin RL, 28. Oh J, Byrd AL, Park M, Program NCS, Kong HH, Segre
Miller JF, Sodergren E, Craft N, Weinstock GM, Li H: Pro- JA: Temporal stability of the human skin microbiome. Cell
pionibacterium acnes strain populations in the human skin 165, 854–866 (2016)
microbiome associated with acne. J Invest Dermatol 133, 29. Rollason J, McDowell A, Albert HB, Barnard E, Worthing-
2152–2160 (2013) ton T, Hilton AC, Vernallis A, Patrick S, Elliott T, Lambert
15. Kilian M, Scholz CF, Lomholt HB: Multilocus sequence P: Genotypic and antimicrobial characterisation of Propi-
typing and phylogenetic analysis of Propionibacterium onibacterium acnes isolates from surgically excised lum-
acnes. J Clin Microbiol 50, 1158–1165 (2012) bar disc herniations. Biomed Res Int 2013, 530382 (2013)
16. Lomholt HB, Kilian M: Population genetic analysis of Pro- 30. Dekio I, Culak R, Misra R, Gaulton T, Fang M, Sakamoto
pionibacterium acnes identifies a subpopulation and epi- M, Ohkuma M, Oshima K, Hattori M, Klenk HP, Rajend-
demic clones associated with acne. PLoS One 5, e12277 ram D, Gharbia SE, Shah HN: Dissecting the taxonomic
(2010) heterogeneity within Propionibacterium acnes: proposal
17. McDowell A, Perry AL, Lambert PA, Patrick S: A new for Propionibacterium acnes subsp. acnes subsp. nov. and
phylogenetic group of Propionibacterium acnes. J Med Propionibacterium acnes subsp. elongatum subsp. nov. Int
Microbiol 57, 218–224 (2008) J Syst Evol Microbiol 65, 4776–4787 (2015)
McDowell A, Gao A, Barnard E, Fink C, Murray PI, Dow- 31. Lomholt HB, Kilian M: Clonality and anatomic distribu-
son CG, Nagy I, Lambert PA, Patrick S: A novel multilocus tion on the skin of antibiotic resistant and sensitive Propi-
sequence typing scheme for the opportunistic pathogen onibacterium acnes. Acta Derm Venereol 94, 534–538
Propionibacterium acnes and characterization of type I (2014)
cell surface-associated antigens. Microbiology 157, 1990– 32. Scholz CF, Bruggemann H, Lomholt HB, Tettelin H, Kilian
2003 (2011) M: Genome stability of Propionibacterium acnes: a com-
18. McDowell A, Barnard E, Nagy I, Gao A, Tomida S, Li H, prehensive study of indels and homopolymeric tracts. Sci
Eady A, Cove J, Nord CE, Patrick S: An expanded multi- Rep 6, 20662 (2016)
locus sequence typing scheme for Propionibacterium 33. Tomida S, Nguyen L, Chiu BH, Liu J, Sodergren E, Wein-
acnes: investigation of ‘pathogenic’, ‘commensal’ and anti- stock GM, Li H: Pan-genome and comparative genome
biotic resistant strains. PLoS One 7, e41480 (2012) analyses of Propionibacterium acnes reveal its genomic
19. Scholz CF, Jensen A, Lomholt HB, Bruggemann H, Kilian diversity in the healthy and diseased human skin micro-
M: A novel high-resolution single locus sequence typing biome. MBio 4, e00003–00013 (2013)
European Journal of Microbiology and Immunology
R. L. W. Petersen et al.: P. acnes type III in macular hypomelanosis 1
Supplementary material
Table S1. Data on eight patients (P) and eight controls (C) included in the study
Sex Age Sex Age
P1 Female 19 C1 Female 28
P2 Female 24 C2 Female 31
P3 Female 18 C3 Female 27
P4 Female 22 C4 Female 26
P5 Female 24 C5 Female 25
P6 Female 31 C6 Female 25
P7 Female 29 C7 Female 26
P8 Female 21 C8 Female 24
Table S2. Treatment and treatment responses in six PMH patients treated with benzoyl peroxide
(BPO) daily washes and/or peroral lymecycline (LC) 300 mg once daily for three months
Patient no. Treatment Outcome
P1 BPO, LC Good response with a few residual lesions
P2 BPO, LC Good response
P3 Irregular BPO, no LC Improved with residual lesions
P4 BPO, LC Good response
P7 Irregular BPO, irregular LC Good response with a few residual lesions
P8 BPO, LC Good response
Table S3. Primers used in this study
Primers for NGS-based SLST
SLST_PA_REV 5´-CCTATCCCCTGTGTGCCTTGGCAGTCTCAG-CCGGCTGGCAAATGAGGCAT-3´
SLST_PA_MID1 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAG-ACGAGTGCGT-CAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID2 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAG-ACGCTCGACA-CAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID3 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGAGACGCACTCCAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID4 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGAGCACTGTAGCAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID5 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGATCAGACACGCAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID6 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGATATCGCGAGCAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID7 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGCGTGTCTCTACAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID8 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGCTCGCGTGTCCAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID10 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGTCTCTATGCGCAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID11 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGTGATACGTCTCAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID13 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGCATAGTAGTGCAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID14 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGCGAGAGATACCAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID15 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGATACGACGTACAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID16 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGTCACGTACTACAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID17 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGCGTCTAGTACCAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID18 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGTCTACGTAGCCAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID19 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGTGTACTACTCCAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID20 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGACGACTACAGCAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID21 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGCGTAGACTAGCAGCGGCGCTGCTAAGAACTT-3
SLST_PA_MID22 5´-CCATCTCATCCCTGCGTGTCTCCGACTCAGTACGAGTATGCAGCGGCGCTGCTAAGAACTT-3
Primers for Sanger-sequencing-based SLST
Sanger_SLST_for 5´-CGCCATCAAGGCACCAACAA-3´
Sanger_SLST_rev 5´-ATATCGGCCCGTATTTGGGC-3´
European Journal of Microbiology and Immunology